Hidden Markov Model Track Identification for Project 8 Phase III

Project 8 uses Cyclotron Radiation Emission Spectroscopy (CRES) and the shape of the tritium beta-decay spectrum to determine the absolute mass scale for the neutrino. The next phase (Phase III) of Project 8 is the Cavity CRES Apparatus, which will demonstrate the scalability of CRES while maintaining high precision for energy reconstruction and low backgrounds. Phase III reconstruction will require measurement of the electron’s frequency, pitch angle, and radial position. The cavity has a single readout channel that records voltage as a function of time, which is then transformed into the frequency domain using a short-time Fourier transform. The resulting signal is displayed as a spectrogram, enabling RF background removal and the visualization of the time evolution of the Power Spectral Density. We will discuss optimizing the track-finding parameters by using a Hidden Markov Model to explore a large parameter space. The Phase III track identification software should improve on previous algorithms’ detection efficiency and energy resolution, and should be capable of fast, real-time processing and of identifying the track’s starting frequency for accurate energy reconstruction in beta decays.